Predicting protein structures efficiently and accurately is a critical challenge for molecular modelers. Whether you are studying protein folding, exploring binding sites, or designing new therapeutics, generating high-quality protein models is often both a necessity and a pain point. With the AlphaFold-2 service on SAMSON’s Biomolecular Structure Prediction extension, you have access to cutting-edge structural prediction tools that can make this task faster and more accessible.
Why Use AlphaFold-2?
AlphaFold-2 has set new benchmarks in protein structure prediction by leveraging deep learning to deliver results that are often closer to experimentally determined structures. But tools like this are notorious for being difficult to set up. Here’s where SAMSON steps in: by combining AlphaFold-2 with an intuitive interface and cloud resources, SAMSON empowers users to predict structures effortlessly.
Step-by-Step Workflow for Using AlphaFold-2 on SAMSON
If you are new to AlphaFold-2 on SAMSON, don’t worry. Follow these simple steps to get started:
- Open the Home > Predict section in SAMSON.
- Select AlphaFold-2 as your prediction service.
- Provide the input sequence files in FASTA format (one or more files).
- Choose your modeling preferences, such as the AlphaFold model (e.g., monomer or multimer) and the sequence alignment database.
- Click Start prediction. That’s it!
The entire workflow is streamlined within SAMSON, ensuring a unified experience for modelers.
The Power of the Cloud
All predictions are carried out using cloud computing powered by Nvidia A100 GPUs. This means you don’t need high-end hardware on your local machine, and you can scale resources depending on your computational needs. The secure cloud connections also ensure that your data is protected during every step of the process.
Easy Visualization of Results
Once the prediction is complete, you can immediately view the results in SAMSON. Navigate to Interface > Cloud jobs in SAMSON or check SAMSON Connect > Account > Jobs. The predicted structures are colorized based on their pLDDT values (if available), giving you a quick visual indication of prediction confidence.
Citations and Acknowledgments
If you publish results using the AlphaFold-2 service in SAMSON, make sure to properly cite the AlphaFold paper. Additionally, include the appropriate references if you choose specific multimer models during analysis.
Credits for Predictions
Using AlphaFold-2 on SAMSON requires computing credits, which can be purchased directly through SAMSON Connect. You can also contact the SAMSON team to request credits if needed.
Start Predicting Today
AlphaFold-2 integration into SAMSON bridges the gap between advanced machine learning tools and user-friendly workflows. To learn more about AlphaFold-2 and its implementation in SAMSON, you can visit the official documentation.
SAMSON and all SAMSON Extensions are free for non-commercial use. Get started today by visiting SAMSON Connect.